Radiotherapy for sinonasal undifferentiated carcinoma

AM ER IC AN JOUR NA L OF OTOLARY NG OLOG Y –H EA D A N D N E CK ME D I CI N E AN D SUR G E RY 3 5 ( 2 0 14 ) 14 1–1 4 6

Available online at www.sciencedirect.com

ScienceDirect www.elsevier.com/locate/amjoto

Radiotherapy for sinonasal undifferentiated carcinoma☆ Kaitlin Christopherson, BS a, c , John W. Werning, MD b , Robert S. Malyapa, MD, PhD a, c , Christopher G. Morris, MS a, c , William M. Mendenhall, MD a, c,⁎ a b c

Department of Radiation Oncology, Gainesville, FL Otolaryngology, University of Florida, College of Medicine, Gainesville, FL University of Florida Proton Therapy Institute, Jacksonville, FL

ARTI CLE I NFO

A BS TRACT

Article history:

Purpose: To evaluate the long-term effectiveness of radiotherapy (RT) in the treatment of

Received 22 July 2013

sinonasal undifferentiated carcinoma (SNUC). Materials and methods: The medical records of 23 patients treated with definitive or postoperative RT between 1992 and 2010 at the University of Florida were retrospectively reviewed. Fifteen patients (65%) received primary surgery and postoperative RT. Radiation doses ranged from 59.0 to 74.8 Gy (median, 70.2 Gy). The median follow-up time for all patients was 3.0 years (range, 0.9–19.9), and for living patients was 7.7 years (range, 2.5–19.9). Results: The actuarial 5-year survival outcomes were as follows: progression-free survival, 42%; cause-specific survival, 43%; and overall survival, 32%. Actuarial 5-year disease control rates were as follows: local control (infield or marginal), 74%; local-regional control (excluding leptomeningeal spread), 58%, regional control 78%, freedom from leptomeningeal recurrence, 72%, and distant metastasis-free survival, 73%. Five of the 8 (62.5%) patients treated with definitive RT died with disease, and 6 of the 15 patients (40%) treated with primary surgery and postoperative RT died with disease. Three patients (13%) experienced severe complications including unilateral eye removal, osteoradionecrosis of the maxilla requiring hyperbaric oxygen and surgery, and brain necrosis. One patient died due to an infected bone graft and brain abscess. Conclusions: A multimodal approach is best when treating SNUC patients. The prognosis for patients treated with definitive RT ± chemotherapy is less promising than for those who receive surgery and postoperative RT ± chemotherapy. Severe complications occur in about 17% of patients due to the high dose of RT alone or combined with surgery required for acceptable disease control. © 2014 Elsevier Inc. All rights reserved.

1.

Introduction

Carcinomas of the nasal cavity and paranasal sinuses are rare, making up less than 1% of all cancers [1,2]. Sinonasal undifferentiated carcinoma (SNUC) was first described less than 30 years ago [3] as an ectodermally derived tumor arising

from Schneiderian epithelium in the nasal cavity or paranasal sinuses [3–5]. It is clinically distinct from other neuroendocrine tumors in the nasal cavity and sinuses, such as esthesioneuroblastoma [6]. SNUC is histologically heterogeneous, with its diagnosis requiring microscopic and immunohistological evaluation [5]. It is also very uncommon, with less

☆

Financial support. None. ⁎ Corresponding author at: Department of Radiation Oncology, 2000 SW Archer Rd., PO Box 100385, Gainesville, FL 32610-0385. Tel.: + 1 352 265 0287; fax: + 1 352 265 0759. E-mail address: [email protected] (W.M. Mendenhall). 0196-0709/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjoto.2013.10.001

142

AM ER IC AN JOUR NA L OF OTOLARY NG OLOG Y –H EA D A N D N E CK ME D I CI NE AN D SUR G E RY 3 5 ( 2 0 14 ) 14 1–1 4 6

than 200 cases reported in the literature [7]. The disease has a high propensity to recur, with reported local-regional recurrence rates ranging from 50% to 63% [8]. In the largest metaanalysis review to date, Reiersen et al. reported on 167 cases identified and found the overall disease-free survival rate to be 26.3%. Patients typically present with nondescript symptoms, including epitaxis, facial swelling or pain, periorbital swelling, diplopia, nasal obstruction, proptosis, cranial nerve deficits, headaches, or a combination of these symptoms [3,5,6,9]. Symptoms tend to develop quickly, typically between weeks and months before the diagnosis of SNUC is entertained [5,6]. At the time of diagnosis, the lesion is usually advanced [10–12], with extension beyond the nasal cavity and paranasal sinuses with involvement of the orbit(s) and/or brain [5]. Some case reviews have observed that 84% to 92% of patients present with T4 disease as defined by the American Joint Committee on Cancer (AJCC) [13,14]. SNUC is most commonly diagnosed in males, with a 2:1 to 3:1 male to female preponderance [6,7,13]. The age range at diagnosis has been reported to be from 12 to 84 years [7], with the mean age in the fifth decade [6–8,12]. SNUC is histologically identified by the overexpression of cytokeratin markers [1,8]. Various studies have attempted to define the etiology of SNUC. Cigarette smoking has been implicated by Frierson et al. [3] and Jeng et al. [10]. There is controversy over whether the Epstein-Barr Virus plays a role in the development of SNUC [4,6,10]. Wadsworth et al. reported over-expression of p16 without the human papillomavirus (HPV) genome in some SNUC cases [15]. SNUC may arise as a second malignancy following radiotherapy (RT) for nasopharyngeal carcinoma [10], or it may present after retinoblastoma [3]. None of these correlations are clearly defined. Because of the rarity of SNUC, its advanced presentation at diagnosis, and its aggressive nature, retrospective reviews continue to be of utmost importance in evaluating and improving treatment protocols. Current treatment typically includes a multimodal approach [16] including RT alone or combined with surgery, with or without adjuvant chemotherapy [4,7,12]. The role of chemotherapy is ill-defined largely due to the rarity of SNUC [7]. Local control relies on high RT doses; thus, the use of proton therapy to produce a more conformal dose distribution may be beneficial in minimizing side effects when available.

2.

Methods

Between 1992 and 2010, 23 patients were treated with RT for SNUC at the University of Florida Department of Radiation Oncology in Gainesville, Florida, and the University of Florida Proton Therapy Institute in Jacksonville, Florida. Radiation oncology medical records were retrospectively reviewed under an institutional review board-approved protocol. Patients included in this study were treated with curative intent, had no distant metastases at presentation, and had the potential for 2-year minimum follow-up. Patients who presented with disease beyond the primary tumor site and cervical lymph nodes or were treated palliatively were excluded. Patient and treatment characteristics are summarized in Table 1. The median age was 56.5 years (range, 23–

Table 1 – Patient and Treatment Characteristics (N = 23). Parameter Male White Radiotherapy Alone Preoperative Postoperative Radiotherapy modality Photons Protons + photons Treated twice daily Chemotherapy ⁎ Induction Concurrent Maintenance None Overall American Joint Committee on Cancer Stage III IVa IVb

No. patients (%) 14 (61%) 19 (83%) 8 (35%) 2 (9%) 13 (56%) 18 (78%) 5 (22%) 20 (87%) 4 14 6 7

(17%) (61%) (26%) (30%)

1 (4%) 7 (30%) 15 (65%)

⁎ Some patients received more than one chemotherapy regimen.

83 years). Median follow-up for all patients was 3.0 years (range, 0.9 to 19.9 years), and for survivors was 7.7 years (range, 5.5 to 19.9 years). No patient was lost to follow-up. Five of the 23 patients (22%) received proton therapy as some portion of their treatment, and 18 (78%) patients received conventional photon therapy only. Due to the high likelihood of cervical lymph node metastases, 13 (72%) out of 18 patients with a clinically negative neck (cNo) received elective neck irradiation (ENI). No patients underwent neck dissection as part of their initial management. All patients with cervical lymph node involvement at presentation received chemotherapy. Ten patients (43%) received trimodality treatment consisting of surgery, RT, and chemotherapy; 5 patients were treated with surgery and postoperative RT; 6 patients were treated with chemoradiation; and 2 patients were treated with definitive RT alone. The median RT dose to the primary tumor was 7020 cGy (range, 5940–7480). The median number of fractions was 54 (range, 39–68). The median fraction size was 120 cGy (range, 110–200 cGy). When the neck nodes were irradiated, the doses ranged from 5000 cGy to 7000 cGy based on clinical presentation. All patients had a continuous course of RT, with no unintended or prescribed breaks. All patients received their prescribed RT dose, except for 1 patient who terminated treatment at 6040 cGy due to severe depression. Twenty patients (87%) received twice-daily fractionation and 3 patients were irradiated once-daily. Sixteen patients (70%) received adjuvant chemotherapy during their initial treatment, including all 5 patients with clinically positive neck nodes. Chemotherapy regimens varied greatly among treatment eras. Many patients received more than one chemotherapy regimen. Weekly concurrent cisplatin was the most commonly employed regimen, used in 10 patients (43%). Six patients received maintenance chemotherapy, 4 received induction chemotherapy, and 14 received concomitant chemotherapy with RT.

AM ER IC AN JOUR NA L OF OTOLARY NG OLOG Y –H EA D A N D N E CK ME D I CI N E AN D SUR G E RY 3 5 ( 2 0 14 ) 14 1–1 4 6

2.1.

Statistical analysis

Statistical analyses were performed using SAS software (SAS Institute, Cary, NC) and the Kaplan-Meier product-limit method to evaluate rates of local control, regional control, local-regional control, distant metastasis-free survival, causespecific survival, and overall survival.

3.

Results

At the time of analysis, 8 patients (35%) were alive; 7 patients had no evidence of disease and 1 patient was alive with disease. Fifteen patients (65%) have died, including 1 patient who died of treatment complications, 11 patients who died with disease, and 3 patients who died of intercurrent disease. SNUC recurred after treatment in 12 patients (52%). The 5-year overall survival, cause-specific survival, and progression-free survival rates were 32%, 43%, and 42%, respectively. Local control, regional control, and distant metastasis rates for various treatment groups are summarized in Table 2. The regional control rates for cN0 patients was 12 of 13 patients (92%) for those receiving ENI versus 3 of 5 (60%) who did not receive ENI. The neck was controlled in 3 of 5 patients (60%) who presented with positive cervical nodes (p = 0.1008). We failed to find a statistically significant difference in the local control rate for patients treated with surgery and RT versus patients treated with RT alone (p = 0.1799). We also failed to find a difference between these groups for causespecific survival (p = 0.5986). Crude ratios show that 62.5% (5/ 8) of patients treated with definitive RT +/− chemotherapy died with disease, while only 40% (6/15) of patients treated with surgery and adjuvant RT died with disease. For patients who received chemotherapy, 31% (5/16) are alive at last follow-up versus 42% (3/7) who did not receive any chemotherapy. Of the patients who received weekly concomitant cisplatin chemotherapy, 50% (5/10) are currently alive, 4 of whom have no evidence of disease. For patients on other

143

chemotherapy regimens, all are deceased: 2 of 6 died of intercurrent disease and 4 of 6 died with disease. The local control rate at 5 years was 74%. There were 5 local failures (22%) that occurred between 0.9 to 3.0 years (median, 1.5 years) after treatment. Four failures were in-field, and one was a marginal local failure. Overall, 37.5% (3/8) of patients treated with definitive RT recurred locally, while only 13% (2/15) of patients treated with surgery + RT recurred locally. Four patients (80%) with local recurrence ultimately died with disease. We separately account for leptomeningeal spread, as this is different than the recurrence of a primary tumor or bloodborne distant metastases. Five patients experienced leptomeningeal recurrence. Freedom from leptomeningeal spread at 5 years was calculated to be 72%. Four patients with leptomeningeal recurrence presented with stage IVb disease and 1 presented with stage IVa disease. Sixty percent of patients with leptomeningeal spread died with disease. One patient, who also had a marginal local failure, is alive with stable dural metastases. One patient was treated with 32.5 Gy for an initial leptomeningeal spread and 30 Gy for a subsequent dural recurrence; the patient was alive as of this analysis at 13 years after treatment. The local–regional control rate (including leptomeningeal spread) at 5 years was 48%. Nine patients failed regionally; the 5-year regional control rate was 58%. Patients with regional failure included 3 of 5 patients with positive regional neck nodes at diagnosis, 2 of 5 patients with a clinically N0 neck who did not receive ENI, and 4 of 13 patients with an N0 neck at presentation who received ENI. Three patients (33%) with regional progression also developed disease progression distantly. Seven of 9 patients with a regional recurrence died with disease. Six patients (26%) developed distant metastases. The 5year metastases-free survival rate for the entire cohort was 73%, and the median time to distant failure was 0.8 years (range, 0.6 to 1.7 years). Five of the 6 patients had overall stage IVb at presentation; 2 of the 6 patients recurred locally. Three of 5 patients (60%) who presented with positive neck nodes

Table 2 – Disease control and distant metastases rates. Variable

Adjuvant chemotherapy Yes (n = 16) No (n = 7) RT strategy Pre-op RT (n = 2) Post-op RT (n = 13) RT alone (n = 8) RT type Photon (n = 18) Proton + photon (n = 5) Stage III (n = 1) IVa (n = 7) IVb (n = 15) RT, radiotherapy.

Local control, no. pts. (%)

Regional control, no. pts. (%)

Leptomeningeal spread/metastases, no. pts. (%)

Distant metastases, no. pts. (%)

12 (75%) 6 (86%)

9 (56%) 5 (71%)

4 (25%) 1 (14%)

6 (38%) 0 (0%)

2 (100%) 11 (85%) 5 (62.5%)

1 (50%) 9 (69%) 4 (50%)

1 (50%) 2 (15%) 2 (25%)

1 (50%) 2 (15%) 3 (37.5%)

15 (83%) 3 (60%)

13 (72%) 1 (20%)

2 (11%) 3 (60%)

4 (22%) 2 (40%)

1 (100%) 5 (71%) 12 (80%)

1 (100%) 3 (43%) 10 (67%)

0 1 (14%) 4 (27%)

0 1 (14%) 5 (33%)

144

AM ER IC AN JOUR NA L OF OTOLARY NG OLOG Y –H EA D A N D N E CK ME D I CI NE AN D SUR G E RY 3 5 ( 2 0 14 ) 14 1–1 4 6

developed distant metastases. Some patients had multiple sites of distant metastases, including bone (N = 3), liver (N = 2), chest (N = 2), and pancreas (N = 1).All patients with distant metastases received chemotherapy during their initial treatment. All patients with distant spread died with disease. Four patients (17%) experienced severe complications defined as those that resulted in loss of unilateral or bilateral vision, hospitalization, surgery, or death. All patients who experienced severe toxicities were treated with RT to doses of 6480 cGy or higher. Three of these patients received postoperative RT, and 1 was treated with definitive RT. Two patients developed radionecrosis and 1 patient had unilateral eye removal. The patient requiring unilateral eye removal had two dural recurrences of SNUC that were treated with palliative RT. The eye removal occurred after the third treatment with RT, and the patient is alive and disease free 13 years after treatment. Another patient treated with surgery postoperative RT died from an infection of the bone graft/flap and brain abscess that had required reoperation. Of the patients treated with protonbased RT, none experienced a severe complication.

4.

Discussion

Patients with SNUC continue to have a poor prognosis despite efforts to intensify and modify treatment. Both our study and a review of the literature demonstrate that many institutions use an aggressive multimodality approach to manage this rare disease [5,7,9,16]. Although the optimal therapeutic approach for SNUC has not been determined, the trend is towards surgical resection, when possible, and adjuvant chemoradiotherapy. The meta-analyses performed by Reiersen et al. found an odds ratio of 2.6 (95% CI, 0.82–7.87) for improved survival for patients treated with surgery plus chemotherapy and/or RT compared to patients treated with surgery alone [7]. Two-year survival rates for patients treated with a trimodality approach have been reported to range from 25% to 67% [12,14,17–19]. Table 3 presents a review of the literature. Al-Mamgani et al. reported on 21 cases of SNUC treated between 1996 and 2010 [20]. On multivariate analysis, a

multimodality treatment approach correlated with superior local control. The authors reported an odds ratio of 55 for patients managed with bimodality versus trimodality treatment (p = 0.003), indicating an increased risk of local failure when only two treatment modalities were used. In our series, the crude local recurrence rate did not show improvement for patients receiving trimodal therapy (3 of 10 patients; 33%) versus treatment with ≤ 2 modalities (2 of 13 patients; 15%). However, given the small sample size and retrospective nature of the study, other factors such as selection bias may have influenced our observations. Musy et al. at the University of Virginia (Charlottesville, VA) reported on 19 patients treated for SNUC [18]. They observed a significant improvement (p = 0.076) in the 2-year overall survival rate for patients treated with surgery (64%) compared with patients treated with definitive RT +/− chemotherapy (25%). These findings are consistent with other studies that advocate the use of a multimodality treatment that includes surgery when feasible [8,11,18,20]. Indeed, all survivors in the largest study of SNUC by Jeng et al. received surgery as part of their initial management [10]. A study from the University of California-San Francisco reported on 21 patients with SNUC treated from 1990 to 2004 [17]. This study did not find a significant difference in local control based on initial treatment approach. Yet when comparing patients treated with varying surgical approaches, gross total resection led to a superior local control rate of 74% compared to 24% in patients who received a subtotal resection (p = 0.001). Although this study could not find a difference in local control or survival between treatment modalities, it does highlight the importance of gross total resection. Our own study failed to find a significant difference in local control or cause-specific survival for patients treated with surgery over those treated with definitive RT. Crude numbers (Table 2) show a trend of improved survival and control with surgery. Nevertheless, definitive RT can achieve local control and cure when surgical resection is unattainable. In the treatment of SNUC, the use of adjuvant chemotherapy warrants further investigation. A study by Rischin et al. of

Table 3 – Literature review.

No. of patients Percent who received surgery Percent irradiated Percent who received chemotherapy Median follow-up (month) Local control rate Regional control rate Disease-free survival rate Overall survival rate

Mourad et al. (2012)

Al-Mamgani et al. (2012)

Chen et al. (2008)

Musy et al. (2002)

Christopherson et al. (present series)

18 83%

21 67%

21 90%

20 55%

23 64%

83% 83%

100% 76%

100% 62%

100% 55%

100% 70%

26

54

36

31

36

72% at 3 y; 56% at 4 y 100% at the time of analysis 65% at 3 y; 52% at 4 y

80% at 5 y 90% at 5 y 64% at 5 y

59% at 5 y 90% at the time of analysis † 64% at 5 years

-

74% at 5 y 78% at 5 y 42% ⁎ at 5 y

50% at 3 y; 48% at 4 y

74% at 5 y

43% at 5 y

47% at 2 y

32% at 5 y

⁎ Progression-free survival. † Distant metastases-free survival.

AM ER IC AN JOUR NA L OF OTOLARY NG OLOG Y –H EA D A N D N E CK ME D I CI N E AN D SUR G E RY 3 5 ( 2 0 14 ) 14 1–1 4 6

10 patients with SNUC investigated the use of induction chemotherapy with concurrent chemoradiotherapy [19]. The authors concluded that induction chemotherapy may improve local-regional control and decrease distant metastases. In their study, 4 patients with advanced T4 disease achieved local control, experienced no distant metastases, and were alive with no evidence of disease at last follow-up with a median potential follow-up of 43 months. In our series, adjuvant chemotherapy did not appear to favorably impact outcome. However, patients who received chemotherapy likely had more advanced disease so that negative selection bias probably influenced our observations. There is no consensus on the optimal RT dose for SNUC. In the study by Rischin et al., patients were treated with doses ranging from 50 to 60 Gy [19]. All 4 patients who received a dose of 60 Gy were alive at last follow-up without local progression. In our analysis, all patients who achieved cause-specific survival were treated with doses above 62.4 Gy. These trends might indicate a dose-response relationship, but again, further investigation of this treatment variable is warranted. Definitive chemoradiation in the presence of locally advanced disease may be considered palliative [18], but our study shows that the use of non-surgical management can sometimes lead to prolonged survival for patients with advanced disease. In our series, 3 patients treated with definitive chemoradiation survived over 3 years: 1 patient died of intercurrent disease at 3.4 years and 2 patients are alive with no evidence of disease at 9.2 and 13 years, respectively. ENI is suggested for patients presenting with a clinically negative neck, because of the high propensity for spread to the cervical lymph nodes [6]. Furthermore, if a patient has a regional neck failure, the prognosis is grim [7]. In their study of patients treated at the University of California, San Francisco, Chen et al. reported that ENI was employed in 79% of patients and that the risk of regional failure was low [17]. Similarly, we observed an improvement in regional control for patients presenting with a cN0 neck who received ENI. Therefore, in the management of SNUC, it is best to treat the cN0 with ENI to avoid a neck failure. An additional point is that the toxicity of ENI is negligible. There is very little in the literature regarding leptomeningeal spread following treatment for sinonasal undifferentiated carcinoma. All 4 patients in our series with leptomeningeal spread presented with stage 4 disease. Treatment of single lesions of the meninges may be salvaged with radiotherapy. In addition to monitoring for distant metastases, patients require follow-up for potential late effects of high-dose RT. We observed severe complications in 4 (22%) of the 18 patients treated with conventional photon-based RT. One patient died of treatment complications, while the other 3 are alive with no evidence of disease. None of the 5 patients treated with proton therapy experienced a severe or fatal complication. Heavyparticle radiation minimizes the dose to nearby critical and sensitive structures, and our early findings suggest that proton therapy may result in a lower probability of severe complications and should be considered for patients with SNUC when possible.

5.

145

Conclusion

Patients with SNUC amenable to a gross total resection should be treated with surgery followed by postoperative RT. RT doses range from approximately 65 Gy to 75 Gy depending on the dose-fractionation schedule and the suspected density of residual subclinical disease. We prefer twice-daily fractionation to reduce the risk of late complications. Patients with incompletely resectable disease receive 74.4 Gy in 62 twicedaily fractions. We recommend concomitant weekly cisplatin 30 mg/M2 to hopefully improve local-control and reduce the risk of distant failure. We prefer proton RT, when possible to reduce the risk of late complications. Patients who present with a cN0 neck receive ENI to approximately 50 Gy; those with clinically positive nodes and resectable disease undergo resection of the primary lesion and a neck dissection followed by postoperative RT. Patients with incompletely resectable disease are treated with RT to the clinically positive nodes to approximately 70 to 75 Gy followed by a neck dissection if the response is incomplete.

Acknowledgments We would like to thank Jessica Kirwan for editing and preparing the manuscript for publication.

REFERENCES

[1] Franchi A, Moroni M, Massi D, et al. Sinonasal undifferentiated carcinoma, nasopharyngeal-type undifferentiated carcinoma, and keratinizing and nonkeratinizing squamous cell carcinoma express different cytokeratin patterns. Am J Surg Pathol 2002;26:1597–604. [2] Khademi B, Moradi A, Hoseini S, et al. Malignant neoplasms of the sinonasal tract: report of 71 patients and literature review and analysis. Oral Maxillofac Surg 2009;13:191–9. [3] Frierson Jr HF, Mills SE, Fechner RE, et al. Sinonasal undifferentiated carcinoma. An aggressive neoplasm derived from schneiderian epithelium and distinct from olfactory neuroblastoma. Am J Surg Pathol 1986;10:771–9. [4] Donald PJ. Sinonasal undifferentiated carcinoma with intracranial extension. Skull Base 2006;16:67–74. [5] Ejaz A, Wenig BM. Sinonasal undifferentiated carcinoma: clinical and pathologic features and a discussion on classification, cellular differentiation, and differential diagnosis. Adv Anat Pathol 2005;12:134–43. [6] Mendenhall WM, Mendenhall CM, Riggs Jr CE, et al. Sinonasal undifferentiated carcinoma. Am J Clin Oncol 2006;29:27–31. [7] Reiersen DA, Pahilan ME, Devaiah AK. Meta-analysis of treatment outcomes for sinonasal undifferentiated carcinoma. Otolaryngol Head Neck Surg 2012;147:7–14. [8] Kim BS, Vongtama R, Juillard G. Sinonasal undifferentiated carcinoma: case series and literature review. Am J Otolaryngol 2004;25:162–6. [9] Pitman KT, Costantino PD, Lassen LF. Sinonasal undifferentiated carcinoma: current trends in treatment. Skull Base Surg 1995;5:269–72. [10] Jeng YM, Sung MT, Fang CL, et al. Sinonasal undifferentiated carcinoma and nasopharyngeal-type undifferentiated

146

[11]

[12] [13]

[14]

[15]

AM ER IC AN JOUR NA L OF OTOLARY NG OLOG Y –H EA D A N D N E CK ME D I CI NE AN D SUR G E RY 3 5 ( 2 0 14 ) 14 1–1 4 6

carcinoma: two clinically, biologically, and histopathologically distinct entities. Am J Surg Pathol 2002;26:371–6. Mourad WF, Hauerstock D, Shourbaji RA, et al. Trimodality management of sinonasal undifferentiated carcinoma and review of the literature. Am J Clin Oncol 2012. [Epub ahead of print]. Tanzler ED, Morris CG, Orlando CA, et al. Management of sinonasal undifferentiated carcinoma. Head Neck 2008;30:595–9. Lin EM, Sparano A, Spalding A, et al. Sinonasal undifferentiated carcinoma: a 13-year experience at a single institution. Skull Base 2010;20:61–7. Revenaugh PC, Seth R, Pavlovich JB, et al. Minimally invasive endoscopic resection of sinonasal undifferentiated carcinoma. Am J Otolaryngol 2011;32:464–9. Wadsworth B, Bumpous JM, Martin AW, et al. Expression of p16 in sinonasal undifferentiated carcinoma (SNUC) without associated human papillomavirus (HPV). Head Neck Pathol 2011;5:349–54.

[16] Gorelick J, Ross D, Marentette L, et al. Sinonasal undifferentiated carcinoma: case series and review of the literature. Neurosurgery 2000;47:750–4 [discussion 4–5]. [17] Chen AM, Daly ME, El-Sayed I, et al. Patterns of failure after combined-modality approaches incorporating radiotherapy for sinonasal undifferentiated carcinoma of the head and neck. Int J Radiat Oncol Biol Phys 2008;70:338–43. [18] Musy PY, Reibel JF, Levine PA. Sinonasal undifferentiated carcinoma: the search for a better outcome. Laryngoscope 2002;112:1450–5. [19] Rischin D, Porceddu S, Peters L, et al. Promising results with chemoradiation in patients with sinonasal undifferentiated carcinoma. Head Neck 2004;26:435–41. [20] Al-Mamgani A, van Rooij P, Mehilal R, et al. Combinedmodality treatment improved outcome in sinonasal undifferentiated carcinoma: single-institutional experience of 21 patients and review of the literature. Eur Arch Otorhinolaryngol 2013;270:293–9.